A user-friendly top-like monitoring
- docker run
docker run \
--rm \
-d \
--name topify \
--pid host \
--network host \
--volume "/var/run/docker.sock:/var/run/docker.sock" \
--volume "/etc/os-release:/etc/os-release:ro" \
--volume "/etc/passwd:/etc/passwd:ro" \
--volume "/etc/group:/etc/group:ro" \
-e "REFRESH_INTERVAL=1000"
pcorbel/topify:v0.1.0
- docker-compose
docker-compose up -d topify
Then open your browser to localhost:5000
Simply add your class in the topify/collectors directory.
The class should inherit from BaseCollector
The data of the collector will be exposed on a Flask endpoint named after the _id attribute.
- topify/collectors/my_collector.py
#!/usr/bin/env python
# -*- coding: utf-8 -*-
from dataclasses import dataclass
from ..common.base_collector import BaseCollector
@dataclass
class MyCollector(BaseCollector): # Inherit from BaseCollector
_id: str = "my_collector" # The endpoint will be named after that. Required
my_custom_value: int = 0 # A value of the collector
def collect(self) -> None: # The collect method that will be called to update the values of the collector. Required
self.my_custom_value += 1
- curl to access the collector value
$ curl -X GET "http://localhost:5000/my_collector" -H "accept: application/json"
{"_id": "my_collector", "my_custom_value": 30}
- Use expandable datatable to show raw data from the collector
- Integrate more psutil.network functions to compute rate per second
- Integrate more psutil.network functions to make an iftop-like view
- Try to reduce footprint by moving from Flask to aiohttp